Llama 3 8B vs Ministral 3 3B API pricing
Updated
Llama 3 8B: USD 0.0925 input, USD 0.095 output per 1M tokens (Featherless). Ministral 3 3B: USD 0.10 input, USD 0.10 output per 1M tokens (Mistral AI and 1 other). At 3 input tokens per output token, Ministral 3 3B costs 7.4% more than Llama 3 8B (USD 0.10 vs USD 0.09312 per 1M).
How do Llama 3 8B and Ministral 3 3B differ on price?
- Cheapest input price: Llama 3 8B USD 0.0925 against USD 0.10 per 1M tokens, so Ministral 3 3B costs 1.08× as much (8% more).
- Cheapest output price: Llama 3 8B USD 0.095 against USD 0.10 per 1M tokens, so Ministral 3 3B costs 1.05× as much (5% more).
- Among the 201 indexed models, Llama 3 8B ranks 20th cheapest and Ministral 3 3B 21st on the 3:1 blend.
- No input:output mix changes the answer: Llama 3 8B is cheaper at every mix before cache discounts.
- Both list cached-input discounts: Llama 3 8B at 1 standard price of 3, from USD 0.074 (Featherless, 20% below that provider's input price); Ministral 3 3B at 1 standard price of 2, from USD 0.01 (Mistral AI, 90% below that provider's input price).
- Ministral 3 3B states up to 262,144 tokens of context, 32× the 8,192 of Llama 3 8B.
- Llama 3 8B has 3 providers with a standard price, 1 more than Ministral 3 3B; 1 provider sells both (Amazon Bedrock).
- On one account, Amazon Bedrock charges USD 0.375 for Llama 3 8B (4× its cheapest price) and USD 0.10 for Ministral 3 3B (its cheapest price) at 3:1, the lowest combined price among the providers that sell both.
- Shopping around matters more for Llama 3 8B: its dearest standard price is 4.5× its cheapest, against 1× for Ministral 3 3B.
- Mistral AI's own API is one of 2 providers tied for the lowest standard price we found for Ministral 3 3B, USD 0.10 at 3:1.
- Llama 3 8B has no batch price on file; Ministral 3 3B has a batch price at 1 provider, from USD 0.05 at 3:1 (Mistral AI).
- For the one-developer coding agent preset, the gap is USD 6,369 a year in Ministral 3 3B's favour (Llama 3 8B USD 582.03 a month, Ministral 3 3B USD 51.30).
- Per character of input, the gap depends on the language: Llama 3 8B costs 1.27× as much as Ministral 3 3B on Hindi text but 18% less than Ministral 3 3B on Chinese, because the two tokenizers split those languages differently.
Llama 3 8B and Ministral 3 3B prices side by side
USD per 1M tokens, standard tier. "Cheapest" is the lowest price at 3 input tokens to 1 output token; each line can come from a different provider.
| Price | Llama 3 8B | Ministral 3 3B |
|---|---|---|
| Model page | Llama 3 8B Instruct Meta · 3 providers | Ministral 3 3B Mistral AI · 2 providers |
| Cheapest standard price, blended 3:1 | USD 0.09312 lower USD 0.0925 in / USD 0.095 out at Featherless Featherless page checked 2026-11-15 | USD 0.10 USD 0.10 in / USD 0.10 out at Mistral AI and 1 other Mistral AI page checked 2026-11-14 |
| Cheapest input | USD 0.0925 lower Featherless source | USD 0.10 Mistral AI source |
| Cheapest cached input | USD 0.074 Featherless source | USD 0.01 lower Mistral AI source |
| Cheapest output | USD 0.095 lower Featherless source | USD 0.10 Mistral AI source |
| Maker's own API | Meta publishes no paid standard price on its own API that we found. | USD 0.10 / USD 0.10 Mistral AI source |
| Median standard price | USD 0.282 / USD 0.60 input / output, 3 prices | USD 0.10 / USD 0.10 input / output, 2 prices |
| Context window | 8,192 tokens, largest stated | 262,144 tokens, largest stated |
| Providers with a standard price | 3 | 2 |
- Lower marks the lower price on a line; tie, equal prices. Other tiers and free rows are on the model pages.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What do Llama 3 8B and Ministral 3 3B cost for a real workload?
30 days at each model's cheapest standard price for the workload (presets of the cost calculator; cached input at the cached price where listed).
| Workload | Llama 3 8B | Ministral 3 3B | Difference |
|---|---|---|---|
| Support chatbot 2,000 requests a day, 1,500 input tokens (60% cached), 300 output | USD 9.04 Featherless | USD 5.94 Mistral AI | USD 3.10 Ministral 3 3B lower |
| RAG over documents 1,000 requests a day, 8,000 input tokens (30% cached), 500 output | USD 80.09 Cloudflare Workers AI | USD 19.02 Mistral AI | USD 61.07 Ministral 3 3B lower |
| Coding agent (one developer) 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning | USD 582.03 Cloudflare Workers AI | USD 51.30 Mistral AI | USD 530.73 Ministral 3 3B lower |
Ministral 3 3B is cheaper on all three workloads. The widest gap is on Coding agent (one developer): the dearer model costs 11× as much as the cheaper one.
Does the cheaper model change with your input:output ratio?
Llama 3 8B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Ministral 3 3B's, so the input:output ratio never changes the answer.
| Input : output tokens | Llama 3 8B | Ministral 3 3B |
|---|---|---|
| 1 : 3 (long answers) | USD 0.09438 lower Featherless | USD 0.10 Mistral AI |
| 1 : 1 | USD 0.09375 lower Featherless | USD 0.10 Mistral AI |
| 3 : 1 (the site’s blend) | USD 0.09312 lower Featherless | USD 0.10 Mistral AI |
| 10 : 1 (retrieval, long prompts) | USD 0.09273 lower Featherless | USD 0.10 Mistral AI |
| 30 : 1 (agents, big contexts) | USD 0.09258 lower Featherless | USD 0.10 Mistral AI |
USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.
What do Llama 3 8B and Ministral 3 3B cost per character in other languages?
Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the Llama 3.3 70B tokenizer for Llama 3 8B and the Ministral 3 14B tokenizer for Ministral 3 3B:
| Content | Llama 3 8B | Ministral 3 3B |
|---|---|---|
| English | USD 0.020 lower Featherless | USD 0.022 Amazon Bedrock |
| Arabic MSA | USD 0.036 Featherless | USD 0.030 lower Amazon Bedrock |
| Egyptian Arabic | USD 0.036 Featherless | USD 0.032 lower Amazon Bedrock |
| Hindi | USD 0.048 Featherless | USD 0.038 lower Amazon Bedrock |
| Chinese | USD 0.074 lower Featherless | USD 0.090 Amazon Bedrock |
| Spanish | USD 0.025 Featherless | USD 0.023 lower Amazon Bedrock |
| Code | USD 0.024 lower Featherless | USD 0.027 Amazon Bedrock |
From the real cost by language dataset (measured tokens per character on FLORES-200 and other texts). Your own text will differ.
Check it on your own traffic
- Llama 3 8B Instruct API pricing: all 3 providers and other tiers.
- Ministral 3 3B API pricing: all 2 providers and other tiers.
- Cost calculator with Llama 3 8B and Ministral 3 3B preloaded
- Llama 3 token counter
- Ministral 3 token counter
- Prompt caching calculator
More comparisons: Llama 3.1 8B Instruct vs Llama 3 8B · Llama 3 70B vs Llama 3 8B · every Llama 3 8B comparison · every Ministral 3 3B comparison · all 662 price comparisons
Questions
Is Llama 3 8B cheaper than Ministral 3 3B?
Yes. On a blend of 3 input tokens to 1 output token, Llama 3 8B costs USD 0.09312 per 1M tokens and Ministral 3 3B USD 0.10, at each model's cheapest standard price. For input alone, Llama 3 8B is cheaper (USD 0.0925 against USD 0.10 per 1M); for output alone, Llama 3 8B is (USD 0.095 against USD 0.10). This compares price only, not quality.
What do Llama 3 8B and Ministral 3 3B cost per 1M tokens?
Llama 3 8B: from USD 0.0925 input and USD 0.095 output at Featherless; median of 3 standard prices USD 0.282 / USD 0.60. Ministral 3 3B: from USD 0.10 input and USD 0.10 output at Mistral AI and 1 other; Mistral AI's own API charges USD 0.10 / USD 0.10; median of 2 standard prices USD 0.10 / USD 0.10. Each price is read from the provider's own pricing page; the source is linked in the table.
Which providers sell both Llama 3 8B and Ministral 3 3B?
1 provider has a standard price for both: Amazon Bedrock (USD 0.375 and USD 0.10 at 3:1). Buying each model from its own cheapest provider costs USD 0.09312 and USD 0.10.
Which is cheaper for a coding agent or a chatbot, Llama 3 8B or Ministral 3 3B?
For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 582.03 a month on Llama 3 8B (Cloudflare Workers AI) against USD 51.30 on Ministral 3 3B (Mistral AI), a difference of USD 530.73. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 9.04 a month on Llama 3 8B (Featherless) against USD 5.94 on Ministral 3 3B (Mistral AI), a difference of USD 3.10. Each uses the cheapest standard price for that workload, cached input included. Try your own numbers.
At what input:output ratio does the cheaper of Llama 3 8B and Ministral 3 3B change?
Llama 3 8B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Ministral 3 3B's, so the input:output ratio never changes the answer.
Do Llama 3 8B and Ministral 3 3B have cached input pricing?
Llama 3 8B from USD 0.074 per 1M cached input tokens (Featherless; 1 of 3 standard prices list a discount). Ministral 3 3B from USD 0.01 per 1M cached input tokens (Mistral AI; 1 of 2 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.
Which has the larger context window, Llama 3 8B or Ministral 3 3B?
Ministral 3 3B: up to 262,144 tokens, against 8,192 for Llama 3 8B, as stated on providers' pricing pages. Some providers offer less than the maximum; check the model pages before you pick the cheapest row for long prompts.